By Don Peppers and Martha Rogers, Ph.D.
A recent study conducted by privacy research guru Dr. Alan F. Westin and Harris Interactive concluded, to nobody's surprise, that the safeguards included in the Health Insurance Portability and Accountability Act (HIPAA) have become something of a roadblock for those hoping to use health data for research purposes. Indeed, the report revealed a significant amount of mistrust, with 58 percent of respondents saying that the privacy of personal health information is not protected well enough.
Given his prominence in the health information/privacy debate -- he holds the titles of Professor of Public Law and Government Emeritus at Columbia University and Director, Health Privacy Program at the Privacy Consulting Group-- Westin's main goal in conducting the research was to get a big-picture take on the public's attitudes about the use of personal health information in health research, especially in the era of electronic health records. At the same time, the study also prompted him to pay closer attention to the societal and marketplace costs of privacy.
What he concludes is that there are significant costs -- especially if, prompted by worries about the sanctity of health data, people hesitate to allow their personal information to be used in research. "One's health condition and status is a major factor in privacy/confidentiality, and a major motivator behind behavior," Westin says. "For mental-health services or sexual-related health conditions in particular, people have a deep hesitation [about sharing their information]."
This tension underscores a larger issue that looms -- the societal and marketplace costs of privacy. Nowhere is that tension more clear than in the recent debates about targeted behavioral marketing of ads tailored to users. The reality is that online advertising is symbiotic with online content. Without the ads, the content that users have come to want and rely on would be less plentiful and reliable. While the debate continues about placing limits on targeted behavioral marketing, one aspect of the discussion should focus on the potential impact on the abundance of free content.
Indeed, Westin points to worries about potential discrimination as a primary reason for that hesitation. "The concern over privacy is very often a concern that somebody who has a health condition will either be denied certain benefits from employers, insurers or the government, or be asked to pay a much higher price for them," he explains. "That alone might prompt somebody to keep all of his information private. If enough people feel that way, the research could be diminished."
Fred Cate, a Professor at the Indiana University School of Law-Bloomington and Director of the Indiana University Center for Applied Cybersecurity Research, describes the market and societal costs of privacy as "huge and getting bigger." By way of example, he points to the movement toward more personal medicine. Testing therapies to treat a specific, narrow group of patients becomes nearly impossible when a large percentage of the population opts out of research due to privacy worries.
"Instead of drugs for everybody, you're seeing them just for a certain subset of people -- just for a certain genotype," Cate explains. "If the total population of people who will use that therapy is 10,000, a study size might be 1,000. If 20 percent of the people say no, that study is doomed. You need the increased participation, which you'll only see if the privacy and security houses are in order."
"The sky's not falling yet, but it will in the future if we don't start thinking about these things," he adds.
So how does Westin propose moving forward? In his mind, what the American public will demand before allowing their health and medical records to be used in research is clear notice as to what kind of research project is being conducted and who's doing the research. The migration of health records to electronic formats could muddy the waters, he notes. "The American healthcare community will want to develop ways to use large data sets for the kind of research that hasn't been possible so far, because no one place had enough medical records that could be massaged and used."
Will there be privacy concerns? Of course, but Westin believes that a consent-based system and explicit communication about privacy will ultimately overcome them. As per his study, consumers tend to trust their providers and academics more than they do any for-profit organization. The promise and delivery of prominent data security will be more essential for the latter group; without it, voluntary participation in health research is pretty much dead in the water.
"When the [personal health] information can be accessed with all the right safeguards from the privacy standpoint, it could lead to breakthroughs in research, and fast," he explains. "What we have to figure out is how we can do breakthrough research without people feeling that they've lost control or that their information is being used for private profit. So the price tag for that, especially for for-profit groups, is going to be the development of those strong privacy practices and protections."
Cate, on the other hand, predicts a move away from consent-based research, which he describes as "kind of unworkable, if you're going to try and look at 500,000 already de-identified [medical] records." He believes that the public and the medical community will eventually decide that it's more important to save lives than to protect health data that is already not personally identifiable.
The key here is to use individual information that has been scrupulously de-identified, including all relevant information about each patient in a study. Then we don't have to choose between research that saves lives but violates privacy and failure to conduct potentially life-saving research to protect privacy. Here's the question: Can putting into place an appropriate system that includes scrupulous privacy safeguards change the public's attitude and instill the trust necessary for people to feel comfortable allowing their health data (but not their names) to be used for research purposes?